Could be the L2 cache, but that would be kinda hard to disable. Or is there a jumper?

There are jumpers to deactivate all those on-board peripherals, you can try that.

PD2JK wrote on 2024-05-16, 20:47:

Could be the L2 cache, but that would be kinda hard to disable. Or is there a jumper?

There are jumpers to deactivate all those on-board peripherals, you can try that.

Thank you, nothing in the manual about a jumper for L2 cache, looks like it's toggled in the BIOS.

That said, you're probably on to something with regard to jumpers.

Here's what is really weird: It reports the 133MHz Pentium as 90MHz and it runs the k6-2 500 at 150MHz...but this board, presumably, is supposed to automatically configure CPU parameters because it doesn't have FSB/Multi/Voltage related jumpers. I'm starting to think the chipset itself may be faulty.

Update:

After an evening of researching every single jumper and disabling every non essential component, I still had made no progress.

Then as a curious last resort, I added a second stick of RAM and... it works! No more freeze at the POST screen, board now attempts to find a boot device, keyboard came to life and able to access BIOS.

...but what does this mean? The manual states the board should work fine with just a single stick in DIMM 1, but in this case, it hangs on launch unless both DIMM slots are occupied.

Any thoughts?

My only thought would be "Welcome to the world of PCChips!" Also, carefully inspect the RAM slots for any bent or damaged (oxidation?) contacts.

Repo Man11 wrote on 2024-05-16, 23:29:

My only thought would be "Welcome to the world of PCChips!" Also, carefully inspect the RAM slots for any bent or damaged (oxidation?) contacts.

🤣 indeed, I bought a job lot of non running systems that were found in the back room of a PC shop that closed in 2000, and stored since then.

All of these systems used PCChips socket 7 boards, the first two had exploding AT power supplies in the first hour of use (not surprising after 25 years idle)

But the biggest headache by far has been configuring these boards.

Turns out this time that a new RAM stick has solved the problem and I’m now running on one DIMM without issue. The funny thing is that the outgoing stick actually looks good, no oxidation and working fine in my 6x86, so…I’ve no idea 😂

Repo Man11 wrote on 2024-05-16, 23:29:

My only thought would be "Welcome to the world of PCChips!"

🤣

I have M571 v.3.2A (Elpina branded) which is similar board and it is very picky about installed RAM modules. At least it works well with modified BIOS for AMD K6-III+/2+ with 83 MHz FSB.

analog_programmer wrote on 2024-05-17, 06:47:

I have M571 v.3.2A (Elpina branded) which is similar board and it is very picky about installed RAM modules. At least it works well with modified BIOS for AMD K6-III+/2+ with 83 MHz FSB.

Appreciate this!

Latest update is that it sometimes works with only one ram stick but always with two.

Go figure 🤣

Repo Man11 wrote on 2024-05-16, 23:29:

My only thought would be "Welcome to the world of PCChips!" Also, carefully inspect the RAM slots for any bent or damaged (oxidation?) contacts.

No fan of PC Chips, but trying to use a PC133 256MB RAM DIMM in an early So7 SDRAM board is asking for it, even if you take a high-end Asus or MSI board. Probably that first DIMM was simply not compatible with the old SiS5598 memory controller.

In any event this chipset can't cache more than 128MB (in WT mode, or half that in WB), so 256MB will actually be slower than 64MB.

These PCChips M571(LMR) motherboards use SiS5598 chipset (renamed to something like "TX Pro II") and according to this link: https://www.anandtech.com/show/72/14 the maximum cached RAM is 128 MB which is better than Intel's 430TX chipset cached RAM limit of only 64 MB. I think, such a chipset cache limitations can be avoided by using modded BIOS with AMD K6-2+/III+ and disabled onboard (level 3) cache.

dionb wrote on 2024-05-17, 17:16:

Repo Man11 wrote on 2024-05-16, 23:29:

My only thought would be "Welcome to the world of PCChips!" Also, carefully inspect the RAM slots for any bent or damaged (oxidation?) contacts.

No fan of PC Chips, but trying to use a PC133 256MB RAM DIMM in an early So7 SDRAM board is asking for it, even if you take a high-end Asus or MSI board. Probably that first DIMM was simply not compatible with the old SiS5598 memory controller.

In any event this chipset can't cache more than 128MB (in WT mode, or half that in WB), so 256MB will actually be slower than 64MB.

I'm spoiled - my TXP4 has had no issues with 256 megabyte PC133 SDRAM modules, so I didn't think about how other motherboards might not be as forgiving of memory rated for higher bus speeds than the board was designed for.

Repo Man11 wrote on 2024-05-17, 20:15:

[...]

I'm spoiled - my TXP4 has had no issues with 256 megabyte PC133 SDRAM modules, so I didn't think about how other motherboards might not be as forgiving of memory rated for higher bus speeds than the board was designed for.

The i430TX has essentially the same memory controller as the i440EX (or ZX), which is the second generation of SDRAM technology and far more stable and compatible than first gen i430VX/Via VPx/SiS 5582/5598 controllers.

That said, the bus speed isn't the issue. Memory speed ratings are nothing more than max speeds; it's like road speed limits: you can always go slower, going faster risks crashes. It's the memory density and chip layout that causes compatibility issues. Just try a DIMM with 16Mx4 or 32Mx4 chips on that TXP4 - it will fail completely, as on any Intel chipset.

dionb wrote on 2024-05-17, 21:51:

Repo Man11 wrote on 2024-05-17, 20:15:

[...]

I'm spoiled - my TXP4 has had no issues with 256 megabyte PC133 SDRAM modules, so I didn't think about how other motherboards might not be as forgiving of memory rated for higher bus speeds than the board was designed for.

The i430TX has essentially the same memory controller as the i440EX (or ZX), which is the second generation of SDRAM technology and far more stable and compatible than first gen i430VX/Via VPx/SiS 5582/5598 controllers.

That said, the bus speed isn't the issue. Memory speed ratings are nothing more than max speeds; it's like road speed limits: you can always go slower, going faster risks crashes. It's the memory density and chip layout that causes compatibility issues. Just try a DIMM with 16Mx4 or 32Mx4 chips on that TXP4 - it will fail completely, as on any Intel chipset.

I have one high density stick of PC133 - a PNY 256 meg that's 8x32 (eight memory chips on one side for total capacity of 256 megabytes). The net capacity when installed in my TXP4 is of course only 128 megabytes, but it otherwise works fine. Not to be contradictory, but that's my experience.

Repo Man11 wrote on 2024-05-17, 22:43:

[...]
The i430TX has essentially the same memory controller as the i440EX (or ZX), which is the second generation of SDRAM technology and far more stable and compatible than first gen i430VX/Via VPx/SiS 5582/5598 controllers.

That said, the bus speed isn't the issue. Memory speed ratings are nothing more than max speeds; it's like road speed limits: you can always go slower, going faster risks crashes. It's the memory density and chip layout that causes compatibility issues. Just try a DIMM with 16Mx4 or 32Mx4 chips on that TXP4 - it will fail completely, as on any Intel chipset.

I have one high density stick of PC133 - a PNY 256 meg that's 8x32 (eight memory chips on one side for total capacity of 256 megabytes). The net capacity when installed in my TXP4 is of course only 128 megabytes, but it otherwise works fine. Not to be contradictory, but that's my experience.
[/quote]
That's a DIMM with 8 32Mx8 chips. I was referring to a DIMM with 16 32Mx4 chips. Different situation.

So in the interest of not starting a new thread about this board, I have one more question if anybody can chime in:

The PC came with its ATX form card attached. It has two USB ports and a PS/2 port marked 'mouse'

The USBs work fine.

But my PS/2 optical mouse doesn't power off it at all. Nothing in the BIOS about PS/2 whatsoever, and all the USB options are enabled in there. Anything to try?

If you're trying to use some non-PCChips ATX form card, check the pinout for the connector here: http://m571.com/m571/franczabkar/ps2mouse.htm - this one is for PCChips M571 non-LMR, but must be the same as the one in the M571LMR.

The board also has dedicated PS/2 mouse connector near the keyboard connector.

analog_programmer wrote on 2024-05-18, 05:12:

If you're trying to use some non-PCChips ATX form card, check the pinout for the connector here: http://m571.com/m571/franczabkar/ps2mouse.htm - this one is for PCChips M571 non-LMR, but must be the same as the one in the M571LMR.

The board also has dedicated PS/2 mouse connector near the keyboard connector.

I'm not sure if it's the original one, but thank you very much, I'll take a look!

This is the same board I'm using in my P233MMX Lego build:
Re: Lego PC - Traditional Damascus Theme

It never froze on me on memory check. I use a single stick 64MB memory stick. For PS2, I don't have a photo of the on-board connection, but you can see that on the back it's just a PS2 header screwed to one of the Serial connector holes in the back plane. USB is unused. PS2 mouse works fine...
And yes, you use the BIOS to control FSB speed and multiplier. Along with cache control, this setup allows fine control of speed. I also just use the onboard VGA as it works great in DOS and has drivers for both Win3.1 and Win95. I carve 4MB of memory for the VGA (hi res true color in Win95).